Abstract
The global phase diagram of a doped Kitaev-Heisenberg model is studied using an $SU(2)$ slave-boson mean-field method.
Near the Kitaev limit, $p$-wave superconducting states which break the time-reversal symmetry are stabilized
as reported by You {\it et al.} [Phys. Rev. B {\bf 86}, 085145 (2012)]
irrespective of the sign of the Kitaev interaction.
By further doping, a $d$-wave superconducting state appears when the Kitaev interaction is antiferromagnetic,
while
another $p$-wave superconducting state appears when the Kitaev interaction is ferromagnetic.
This $p$-wave superconducting state does not break the time-reversal symmetry as reported by
Hyart {\it et al.} [Phys. Rev. B {\bf 85}, 140510 (2012)], and such a superconducting state
also appears when the antiferromagnetic Kitaev interaction and the ferromagnetic Heisenberg interaction compete.
This work, thus,
demonstrates the clear difference between the antiferromagnetic Kitaev model and the ferromagnetic Kitaev model
when carriers are doped while these models are equivalent in the undoped limit,
and how novel superconducting states emerge when the Kitaev interaction and the Heisenberg interaction compete.
